How to find out infiltration rate for infiltration trench

[Afrida_09]

Dear Jirka,

I am simulating an infiltration trench that has a height of 0.8 m, top width 1.6 m, bottom width 0.8 m and water level in the trench is 0.4 m. The GWT is below 4.5 from the surface. I considered a saturated zone of 10 m. My goal is to find out the infiltration rate of the trench and the optimal distance between the trenches. I set up the model using symmetric condition.

For determining the infiltration rate I was looking into cum. constant flux boundary. But I think it is the sum of all the constant bc which I have used in the model. Could you please guide me for estimating the infiltration rate where should I look at?

Another question is for the infiltration trench for defining the water level of 0.4 m I only defined it at the bottom of the trench. Is it fine? Or do I need to define it on the side also? I have attached the model here.

[Jirka]

a. Indeed, Hydrus will integrate fluxes across all boundary with the same type of BC. You should use different types of BCs (e.g., constant head, time-variable head 1, 2, 3, etc) if you want the fluxes for different parts of the boundary.
b. You should give the head BC on both the bottom and sides (linearly decreasing to zero; use the equilibrium condition for constant head BC; for time-variable BC, the hydrostatic equilibrium is assigned automatically) of the trench.

J.

[Afrida_09]

Hello Prof. Jirka,

Thank you very much for explaining.
So I applied time var. BC on both sides of the trench, I am confused in some issues-
1) I kept it 0.4 m head (in time var. BC window) for 92 days. I have not applied time var. BC before. Is it correct to do it like this?
2) After providing the BCs and initial condition before running the model I get this message " The following properties have different values". See Pic 1. Did I define any condition wrong?
The updated model is attached here.
3) While I check the var flux in result why there is an increase in flux at the beginning of the simulation, which indicated to high infiltration at the start and then it becomes constant? I hope I am interpreting it right. Should the graph look like this at the initial time steps?

[Jirka]

a. You can use time-variable BC and keep the value constant with time.
b. Pic 1: I have answered this question many times in this forum and do not know what I can add to it.
c. Would not you expect to have a decreasing infiltration flux for a constant head BC? Does not every infiltration equation (Kostiakov, Horton, Philip, etc) have such shape?
J.

[Jirka]

a. You can use time-variable BC and keep the value constant with time.
b. Pic 1: I have answered this question many times in this forum and do not know what I can add to it.
c. Would not you expect to have a decreasing infiltration flux for a constant head BC? Does not every infiltration equation (Kostiakov, Horton, Philip, etc) have such shape?
J.

[Afrida_09]

Thank you Prof. Jirka for your answer.
Normally the infiltration rate when we represent it is in positive value. As hydrus represents the flux value in negative format. Do I need to do anything to represent the value in a positive form? If I just represent the negative value in positive form the graph is in opposite shape (which is expected). In this case what do you suggest?

[Jirka]

In multidimensional modeling, it is a common standard to assume that fluxes directed out of the domain are positive and fluxes directed into the domain are negative. The HYDRUS model follows this convection. You can obviously redo the graphs with an opposite sign. It is just your choice. J.